PACIFIC HERRING 
229 
might make it possible to foretell coming years of scarcity or abundance. Moreover, 
the value of such information to the industry in indicating the extent of preparations 
needed for the coming season would be very great. To determine this age or size 
composition we have , sought to make our samples represent the commercial take ; 
but in a fishery so scattered and in which the season in a given locality may be 
exceedingly short, it is not only difficult to obtain a full representation of the com- 
mercial catch, but it is difficult to be certain that the commercial catch from year to 
year is taken from the same section of the actually existing population. 
In pursuing this study it should be remembered that such knowledge must be 
made use of to explain the fiuctuations in the catch, and that without adequate 
records of the catch from year to year, it may be impossible to establish and prove 
a definite connection between dominant year classes and unusual abundance or 
periods of scarcity. Until such a connection is proved to exist, and until its extent 
can be tested, no prophesies as to the catch can be ventured upon, and it is impossible 
to assume that depletion is not occurring or that the presence of dominant year 
classes in our samples is reflected in the catch. The trade alone can furnish such 
statistics to the Government. 
Since the presence or absence of depletion must manifest itself through the 
commercial take, as evinced by the total catch or by the catch per unit of gear, reliable 
statistics must be obtained. At present the presumption in the public mind is that 
depletion occurs whenever herring become scarce in a given locality; and there is 
no logical way to disprove this save by advancing and proving some other explanation 
for a decline, such as the passage and disappearance of a dominant year class. 
In determining the causes of the fluctuations in abundance one of the most 
important questions to be faced is that of the degree of migration, as upon that 
depends the relative interdependence of the populations of different regions. The 
existence of a single stock of herring, freely intermingling and migrating along tli8 
narrow coastal banks, would mean that any fluctuations or depletion would be wide- 
spread, and that any regulations, to be effective, would have to consider the whole 
coast as a unit. On the other hand, if local “races” were present, each locality would 
have to be treated as a separate unit, since it would then be possible to greatly reduce 
the supply in one area without affecting it elsewhere. A great deal thus depends on 
the existence or nonexistence of local “races” or populations of herring, and much of our 
study has been on this problem. 
Whether any of the changes in abundance can be due to overfishing is a question 
often asked. At present our chief criterion of depletion is afforded by the statistics 
of the commercial catch. In some cases these statistics do not give an adequate or 
detailed enough picture of what has occurred, yet in a few cases the changes in 
abundance have been so great that they would be difficult, indeed, to obscure. How- 
ever, even when a decrease in abundance is shown by statistical criteria, the con- 
clusions drawn from them should be corroborated, if possible, by biological evidence, 
such as a decrease in the abundance of the older fish, a failure in areas or with types 
of gear depending chiefly on the schools of larger fish, or a shift in the fishing grounds. 
The significance of the last-named criterion will depend on the proof of the existence 
of separate stocks of herring in the localities in question. 
Racial data have been gathered on the spawning populations, inasmuch as a 
comparison of the spawning populations with those comprising the main commercial 
catch (taken in the summer months) will show on which spawning areas each of the 
